The use of supplemental oxygen in preterm infants continues to be a challenge as sustained low and high oxygen, and fluctuations in oxygen have been implicated in infant morbidity. In a large multi-center trial (Surfactant, Positive Pressure, and Oxygenation Randomized Trial or SUPPORT), preterm infants randomized to a low level of oxygen revealed an unexpected increase in infant mortality. In addition, data in a subcohort of infants enrolled in the trial showed a higher incidence of fluctuations in oxygen levels, or intermittent hypoxia (IH) events, in the low oxygen saturation target group. Therefore, although the cause for the increased infant mortality is unknown, intermittent hypoxia may have played a key role. Our objective for this proposal is to quantify patterns (incidence, duration, severity an timing) of intermittent hypoxia and their interaction with baseline oxygen saturation in infants who died compared to infants who survived. Existing continuous pulse oximetry recordings from the SUPPORT trial will be reviewed for infants who died and compared to SUPPORT trial infants who survived. A novel computer program and techniques developed by our group will be used to analyze both baseline oxygen saturation and the incidence and patterns of oxygen fluctuations on a daily basis to identify variations in oxygenation associated with infant mortalit. Early identification of risk patterns of oxygen fluctuations may lead to timely intervention to reduce infant mortality in this high risk cohort.